Prior to the present invention, commercially available plastic encapsulants for electronic devices had several serious disadvantages. One disadvantage, for example, was that prior to use, the encapsulant required refrigeration, preferably to 4.degree. C. and protection from moisture during shipment and storage. Unless the encapsulant composition was refrigerated and protected from moisture, it did not provide suitable flow length for filling multicavity molds. Encapsulant molding powders are generally sealed inside several plastic bags and surrounded by dry ice before shipment by the manufacturer. As a result, the material must be allowed several hours to reach room temperature in the absence of moisture before it is used to encapsulate an electronic device to minimize a build-up of water in the powder due to atmospheric condensation.
Although flow length during injection molding is an important consideration for qualifying commercially available curable polymeric materials for device encapsulation, another equally important requirement of the electronic system manufacturers is that after cure, the plastic encapsulated electronic device must have the ability to resist changes in humidity conditions over a wide temperature range. One way to test the resistance of plastic encapsulated electronic devices to high humidity over various temperature ranges, such as experienced in South East Asia, is by using the "HAST" test (highly accelerated stress test) as discussed by L. Gallace et al. in "Reliability of Plastic-Encapsulated Integrated Circuits and Moisture Environments", RCA Solid State Division, Somerville, N.J. 08876; and RCA Review. Vol. 45. (June 1984) pages 249-277. Another version of the HAST test is shown by K. Ogawa et al., Automatically Controlled 2-Vessel Pressure-Cooker Test--Equipment IEEE Transactions on Reliability, Vol. R-32, No. 2 (June 1983). The HAST test determines how long terminals of a plastic encapsulated device will survive while subjected to 18 volts during exposure to 85RH, 145.degree. C. and 2.7 atmospheres of steam. The devices are periodically tested with a GenRad model 1731M/linear IC tester having internal diagnostics and system calibration.
An additional consideration which must be addressed by encapsulant composition manufacturers for the electronic industry is the cure temperature required for commercially available encapsulants which are usually epoxy resin formulations. It is generally known that presently available encapsulants which are injection molded onto electronic devices require a temperature of at least 180.degree. C. Such temperatures often create excessive stresses at the interface of the cured plastic encapsulant and the electronic device upon cooling.
In order to minimize the cracking of silicon chips resulting from 180.degree. C. plastic encapsulation, silicone resin modified epoxy resins have been used as encapsulants, as shown by K. Kuwata et al., IEEE, 1985 (18-22).
An additional concern of the electronic systems manufacturers is that plastic encapsulated electronic devices requiring soldering often fail because the heat distortion temperature (HDT) of the cured encapsulant is often too low, i.e. between about 200.degree. C. to 250.degree. C. The molded encapsulant can experience a change in shape (distortion), when the device is dipped into a molten solder bath. Temperatures up to 290.degree. C. are sometimes unavoidable in instances where solder fluxes which are often used to facilitate metal contact can not be tolerated.
It would be desirable, therefore, to provide less stressful encapsulating compositions curable at temperatures significantly below 180.degree. C., which have HDTs after cure exceeding 300.degree. C. Encapsulated devices fabricated from such encapsulants would have substantially reduced stress due to decreased thermal expansion. In addition, such encapsulated devices could be made with surface areas exceeding 1/8 sq. in., without cracking. It also would be desirable to have encapsulated electronic devices having greater than a 50% survival rate when subjected to HAST test conditions measuring reliability in moisture environments.
The present invention is based on our discovery that if a catalytic amount of a compatible monomeric or polymeric diaryliodonium hexafluoroantimonate salt as defined hereinafter can be utilized in combination with an effective amount of a copper compound, as defined hereinafter, as a cocatalyst, and the resulting thermal initiator is employed in combination with a substantially chloride-free epoxy resin and a fused silica filler, superior silicon chip encapsulating compositions are provided. Many of the aforementioned requirements of the electronic systems industry can be satisfied. For example, the aforementioned diaryliodonium hexafluoroantimonate catalyst and copper compound cocatalyst has been found to provide encapsulating compositions which are shelf stable indefinitely at ambient temperatures and are curable at 150.degree. C. These encapsulants can provide sealed electronic devices having HDTs from about 250.degree. C. to 345.degree. C., and reduced incidence of cracking. The encapsulated devices also have been found to far exceed the 50% survival rate after a period of 80 hours under the aforedescribed HAST test conditions. In addition, the dynamic mechanical properties of the cured encapsulating composition of the present invention, in accordance with ASTM test D2236-81, show that devices encapsulated in accordance with the present invention have reduced likelihood of failure due to shape alteration during device performance as compared to commercially available devices encapsulated with plastic.